HyperAI초신경

홈 뉴스 최신 연구 논문 튜토리얼 데이터셋 백과사전 SOTA LLM 모델 GPU 랭킹 컨퍼런스

한국어

HyperAI초신경

Interactive Segmentation On Sbd

평가 지표

NoC@85

NoC@90

평가 결과

이 벤치마크에서 각 모델의 성능 결과

모델 이름	NoC@85	NoC@90	Paper Title	Repository
RITM (HRNet18, SBD)	3.39	5.43	Reviving Iterative Training with Mask Guidance for Interactive Segmentation
RIS-Net	6.03	-	Regional Interactive Image Segmentation Networks	-
ViT-B+MST+CL	3.03	-	MST: Adaptive Multi-Scale Tokens Guided Interactive Segmentation
UCP-Net	2.73	5.00	UCP-Net: Unstructured Contour Points for Instance Segmentation	-
IA-FP-Net(HRNet,SBD)	3.33	5.25	Cascaded Sparse Feature Propagation Network for Interactive Segmentation
DOS with GC	9.22	12.80	Deep Interactive Object Selection
DOS w/o GC	14.30	16.79	Deep Interactive Object Selection
FocalClick	3.53	5.59	FocalClick: Towards Practical Interactive Image Segmentation
FocusCut	3.40	5.31	FocusCut: Diving Into a Focus View in Interactive Segmentation
Latent diversity	7.41	10.78	Interactive Image Segmentation With Latent Diversity	-
SimpleClick	2.51	4.15	SimpleClick: Interactive Image Segmentation with Simple Vision Transformers
BRS	6.59	9.78	Interactive Image Segmentation via Backpropagating Refinement Scheme	-
f-BRS-B (ResNet-101)	4.81	7.73	f-BRS: Rethinking Backpropagating Refinement for Interactive Segmentation
SimpleClick CFR-1 (ViT-H, SBD)	2.45	4.08	CFR-ICL: Cascade-Forward Refinement with Iterative Click Loss for Interactive Image Segmentation

0 of 14 row(s) selected.